I was thinking about what the percentages in the OE *really* tell us, at different stages of the opening... In my view, the choice for a move should *never* depend on percentages, presented in the OE, except when still at the stage with more than say 250 games left, prior to that move... But a bit later on, if after EG move X has a 37% winning percentage, yet only played at 2400+ level say 27 times, then what can be deducted from those figures, alone?

I'll rephrase:

How many games, that have been won by Black from that position, (10 from 27 in this fictive case), have been decided by a solid building on a realistic, describable, opening advantage? And subsequent winning of those games versus strong opposition (no blunders)?

Very little, maybe even NONE, as most games evolve with evals swinging several times in favor of either player, especially in zeitnot phases...

We can now switch to a move at a stage where still 117 games are played from a certain opening... My question is: What percentage of the, say, 41 Black wins, can be directly linked to the opening advantage, given subsequent strong play, with no big blunders from either side?

OE --- Oh Eeee. Or <Aieeee!> as certain comicbook characters say.

The stats are untrustworthy, except when there are large sample numbers *and* a very clear black/white win difference.

I know a line in the French that starts with a semi-plausible move by black -- actually a losing blunder, a move that should never be played. But OE gives it an OK score, because the CG database has a few games where White blundered and is missing several games where white won.

The other key factor is date and time. Especially in sharp, trendy lines -- Dragon, Benoni, etc -- where a move can be played 500 times before somebody finds a refutation. After that, one side wins almost all the games.

I don't think it's even possible to generalize on the subject. Too many variables. OE can even be useful sometimes, as long as you watch the sample size and check for recent games in the relevant line.

Your question <What percentage of the, say, 41 Black wins, can be directly linked to the opening advantage, given subsequent strong play, with no big blunders from either side?> reminds me of the difference between good and bad books about openings (the printed kind, not engine books).

Quick, cheap opening books annotate by result. If white wins - and no gigantic blunder is apparent - then, ergo, white had the advantage in the opening. But superior books - like Psakhis on the French - can cite games where black was better on move 15, even if he subsequently lost.

Again, though, it's hard to generalise. Psakhis is a GM and a French expert -- he can see past the result and grasp the dynamics of the opening/middlegame transition.

I may be wrong, but I don't think either engines or untitled players have this ability. Of course we can point to dynamic strengths and weaknesses - and the abaci can count their plies - but it doesn't add up to the same thing. The playing strength of the World team, I think, can be attributed to gestalt and distributed processing - and, of course, the forum system.

It is also important to remember that almost all OTB games - no matter how exalted the GMs - are flawed, with errors that even engines can spot (hence those 'swinging evals'). A CC GM will crush an OTB GM: the genius of Kasparov hung on the amount of stuff - opening lines, endings, calculations - that he had in memory or could do in his head.

This also relates to a point I made before: an opening repertoire is essentially an aide-memoire - a tool for OTB play. In CC it can be a disadvantage: one should be able to play any type of position.

Oh, and you will meet a tall, dark ... King's Indian.

<Dom> Exactly! That's why, when investigating a line/system/opening/move, I try and go through the poule of say 60 games or less - and play through every move in every game that I have carefully pre-selected from the DB... With the engine purring along for tasty blunderbees reaching at least 15 or 16 PLY.

Yes, your answer was satisfactory; in fact I pretty much agree with you on all counts.

(Although you didn't give ne an estimate on the percentage of games that ARE likely to be won out of the opening, after engine aided in depth analysis. -- But that is just a time consuming experiment -- but I was wondering what number you'd be tempted to attach to it...)

<I don't think it's even possible to generalize on the subject. Too many variables. OE can even be useful sometimes, as long as you watch the sample size and check for recent games in the relevant line.> That's what I have been doing for a while now, albeit more intuitively, but since a few months ago a more cognitively as well. You are right about the number of variables involved in the percentages. My opinion is that several experienced World team players do not realize these things fully, or they don't show it -- in any case I have seen little discussion on this specific, yet important issue - not enough, but I haven't read every single kibitz.

(which are really all variations on a single equation) seem to actually be correct. Or correct to at least 30 digits, which goes way past the approximations I mentioned earlier. Now I must work out why ...

I know what you mean about players leaning on OE too hard, but most of those with engines should also have access to engine databases, and anyone can check NICbase and the other online databases -- generally more comprehensive than CG, and often searchable in different ways.

Okay, they don't do those neat little white-black-grey diagrams, but anyone using those needs their head examined. As does anyone who thinks, say, that a score of 41.2% out of 37 games is better than a score of 39.3% out of 829 games.

"Approximating Tablebaseness"

Wether the team will have potential to use its super(ior) engine and analysis power will hinge on the choices made over the next 5-7 moves, IMO...

Some eagerness seems to have gone out of the team now, in comparison... But maybe I am wrong. (I hope)

You are right on the money. And I think it is a very good team project.

GM Shabalov in the September 2008 Chess Life magazine reviews the 15th edition of the MCO, but goes right to chess engines (page 39) in his article.

He writes, "It is important to run at least two engines simultaneously to get a good grip on the computers' analysis. Sometimes Rybka gets over-optimistic and Fritz is very good catching it. On other occasions, when Fritz becomes too cautious, Rybka looks like a creative genius."

Please add any other ideas to the Fritz mix and thanks again!

literally: sound colours

There indeed seems to be a direct link with <synaestthesia>; the (involuntary) ability to experience certain colours following sounds, shapes and smells, has been described in detail.

It simply poses the question:

What do Satan and Beélzebub have to do with Fibonacci's 'golden cut'?

'snede' can have several meanings, among which a figurative one, in the sense of "revealing [laying bare] a solution."

Well please give a detailed report of the obscure <Greenaway> film <Mrs. Wagstaff's Suticases>?

they don't post his movies on the Net.

Also, <Anxiety of the Goal Keeper Upon Facing the Penalty Kick>

heh--

<Wim Wenders> manages to make an entire film about the Holocaust without ever mentioning it.

I like the flourishes-- the camera will almost imperceptibly zoom in on a furnace grate while the characters are talking-- and you think "what's this closeup for?"

Also the frequent montage where he cuts away to shots of passenger trains entering and leaving stations-- shots that have no purpose in advancing the "plot" of the film.

Very clever stuff.

Here is my sixteenth millionth post in your forum!!

Hmm... I don't really have anything to say though...

Which kind of begs the question, "Why did I make sixteen million posts in the first place"?

Under scratch your head cinema I offer The Discrete Charm of the Bourgeoisie by the great Luis Bunuel. I had to see it three times before I finally understood the anxiety I felt. It's all about food but nobody manages so much as a morsel until Fernando Rey is caught under the dinner table with a turkey leg in his mouth. He is then machine gunned by a terrorist.

I've just been looking at an actually new-ish (2002) German film: seems gut. The English title is <What to do in case of Fire> ... nostalgia for 80s punk squatter anarchists rather than the usual '68 revolutionaries ...

Gottfried H was wryly funny (Oh really? No, O'Reilly, oh *wryly*) on such topics, eg

"There were Trotskyites and Maoists and they were always fighting with each other. They came from the middle and upper classes but they talked about the workers - and the real workers thought they were weirdoes and idiots ..."